metrics
HEART framework
What is Google's HEART framework for measuring UX?
HEART is Google's framework for measuring user experience across five dimensions - Happiness, Engagement, Adoption, Retention and Task success - so you capture quality of experience, not just business output. Its real engine is the "Goals-Signals-Metrics" process that turns each fuzzy dimension into a concrete number, forcing you to decide what success means before you pick what to count.
Also known as: heart framework, heart metrics, happiness engagement adoption retention task success
Prefer to watch?
Watch the recap 0:54
The demo
“Good UX” is hard to measure - until you break it into five angles. Click each letter of HEART to see what it captures, and how a fuzzy goal becomes a real metric.
What this demo shows (text version)
The HEART framework's five dimensions are shown as clickable letters. Happiness measures attitudes (satisfaction, perceived ease, NPS). Engagement measures depth of involvement (frequency and intensity of use). Adoption measures new users or uptake of a feature. Retention measures whether people keep coming back. Task success measures efficiency and effectiveness (completion rate, time on task, error rate). Selecting one shows an example Goal → Signal → Metric for it.
The point is two-fold: HEART gives a balanced view of experience across several dimensions so you don't optimise one number while another degrades, and the Goals-Signals-Metrics process forces you to define success (goal), decide what behaviour reveals it (signal), and only then choose what to count (metric). You pick the dimensions that matter for the project rather than tracking all five, and pair it with a north star and counter-metrics.
HEART gives UX a balanced scorecard - Happiness, Engagement, Adoption, Retention, Task success - so you measure the experience from several angles instead of one convenient number. The discipline that makes it work is Goals → Signals → Metrics: state the goal for a dimension, decide what user behaviour would signal it, then pick the metric. You won't use all five for everything; choose the dimensions that matter for the project.
Click through the five and "good UX" stops being a vibe and becomes measurable from five angles - are people happy, engaging, adopting, sticking around, succeeding at their task? Each turns into a concrete signal and metric. That's HEART's gift: it makes a fuzzy experience into something you can actually track without collapsing it to a single misleading figure.
The five dimensions cover different facets: Happiness (attitudes - satisfaction, perceived ease, NPS), Engagement (depth of involvement - frequency, intensity), Adoption (new users or uptake of a feature), Retention (whether people keep coming back), and Task success (efficiency and effectiveness - completion rate, time on task, errors). Together they resist the trap of optimising one number while another quietly degrades.
The part teams skip - and shouldn't - is Goals-Signals-Metrics. For each dimension you care about, first state the goal ("users find what they're searching for"), then the signal (the behaviour that reveals it - they click a result and don't immediately bounce back), then the metric (search success rate). Doing it in that order stops you measuring whatever is easy and calling it success.
HEART is a menu, not a mandate. You rarely track all five; you pick the dimensions that matter for the feature and stage - a new feature leans on Adoption and Task success, a mature one on Retention and Happiness. It pairs naturally with a north star (the headline) and counter-metrics, giving you a balanced set rather than a single number to game.